AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Deep Networks with Stochastic Depth

Gao Huang, Yu Sun, Zhuang Liu, Daniel Sedra, Kilian Q. Weinberger

ECCV 2016 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 2,247 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add stochastic depthResNet-110 · CIFAR-10+test error (%): 6.41 -> 5.25Improved
add stochastic depthResNet-110 · CIFAR-100+test error (%): 27.22 -> 24.98Improved
add stochastic depthResNet-110 · CIFAR-10 (no augmentation)test error (%): 13.63 -> 11.66Improved
add stochastic depthResNet-110 · CIFAR-100 (no augmentation)test error (%): 44.74 -> 37.8Improved
add stochastic depthResNet-152 · SVHNtest error (%): 1.80 -> 1.75Improved
add stochastic depthResNet-1202 · CIFAR-10test error (%): 6.67 -> 4.91Improved
add stochastic depth, 90 epochsResNet-152 · ImageNet valtop-1 error (%): 23.06 -> 23.38Hurt
add stochastic depth with 120 epochs vs baseline 120 epochsResNet-152 · ImageNet valtop-1 error (%): 21.78 -> 21.98No clear effect
increase depth from 110 to 1202 layers with stochastic depth (ablation)ResNet · CIFAR-10test error (%): 5.25 -> 4.91Improved
increase depth from 110 to 1202 layers without stochastic depth (ablation)ResNet constant depth · CIFAR-10test error (%): 6.41 -> 6.67Hurt
use linear decay instead of uniform survival probability (ablation)stochastic depth · CIFAR-10test error (%)Improved
stochastic depth with p_L=0.2 (ablation)stochastic depth linear decay · CIFAR-10training time: 40% reduction in training timeImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy